Speaker
Agnieszka Czaplicka, IFISC

Description

We consider the competition of two mechanisms for adoption processes: a so-called complex threshold dynamics and a simple Susceptible-Infected-Susceptible (SIS) model. Separately, these mechanisms lead, respectively, to first order and continuous transitions between non-adoption and adoption phases. We couple the two adoption processes in a complex network with two interconnected layers. We find that the transition points and also the nature of the transitions are modified in the coupled dynamics. In the complex adoption layer, the critical threshold required for extension of adoption increases with interlayer connectivity whereas in the case of an isolated single network it would decrease with average connectivity. In addition, the transition can become continuous depending on the detailed inter and intralayer connectivities. In the SIS layer, any interlayer connectivity leads to the extension of the adopter phase. Besides, a new transition appears as as sudden drop of the fraction of adopters in the SIS layer. The main numerical findings are described by a mean-field type analytical approach appropriately developed for the threshold-SIS coupled system.

Speaker
David Miralles, La Salle - Universitat Ramon Llull

Description

Interfaces allow us to interact with objects or systems. Technology has changed how we understand interface design. Tangible interfaces, internet of things, and augmented reality are a few of them that invite us to rethink our relationship with objects. Now, machine learning techniques are being used within this field. Objects are beginning to decide for us.  Can it possible to make interfaces disappear? Could they be made seamless?

Speaker
Fakhteh Ghanbarnejad, Technische Universität Berlin

Description

Susceptible-Infective-Susceptible (SIS) and Susceptible-Infective-Recovered (SIR) are two successful models for understanding the dynamics of infectious diseases. Nevertheless deadly cases like tuberculosis during the 1918-1919 Spanish Flu and unexpected HIV cases in presence of hepatitis B and C, TB and Malaria, and vice versa, showed us that interaction between two spreading dynamics can dramatically change the epidemic dynamics. Here we investigate the mechanisms which might lead to the unexpected outbreaks; In this work, we study spreading of two cooperative diseases: a SIS and a SIR dynamics and address similarities and differences in comparison to other minimal cooperative models, i.e. SIR-SIR [1] and SIS-SIS [2]. We build up an idealized and simplified model and treat it in mean field approximations as well as stochastic agent based models. Thus in a mean field treatment we calculate the cooperative epidemic threshold and also outbreak size accordingly. We find out in presence of cooperation an emerging region in the parameter space where the stable endemic and stable free-disease states co-exist. Interestingly this region appears differently in comparison to the SIR-SIR and SIS-SIS models. Also We track the dynamics on random generated networks; And argue how topological features can facilitate or neutralize the cooperation effect.

[1] EPL 104 (2013) 50001; Nature Physics 11, 936–940 (2015); Rev. E 93, 042316.
[2] arXiv:1603.09082v1.

Description

This activity is aimed at the researcher who has probably taken an introductory statistics and probability course at some stage and would like a brief introduction to the core methods of statistics and how they are applied. We are going to do a panoramic view by the good practices in scientific methodologies from statistical point of view. 

The outline of course is over the Statistics as the base of the based on experience scientific method. The first ingredient are data and the second are the models. The third is how to quantify the link between them, called likelihood.

Table of contents

Part 1  -    ​Datasets and Random Variables
Data can have different nature and complexity, according to the way the are obtained and the way they are displayed (multidimansional data, functional data, dependent data,...). We shall review the process of data obtention and several ways to give them structure, according to their nature, their (possible) structures of dependence, based on the primigenial objective of the experiment that produced them. We shall review the probabilistic foundations of the main data reduction techniques and statistical analysis.

Part 2  -    ​Statistical Models and Probability theory
​In order to capture the essense of statistical modeling, we will begin with simple models. We then will add complexity in order to show a large amount of models, from an independent setting (parametric and non-parametric models) to learining models based on data mining techniques (neural network, random forest, ...) passing through classical regression and classification models (mixed model, generalized linear model, …) as well as temporal dependence (autoregressive, moving average,…). We will review these models from a probabilistic point of view.

Part 3  -    ​Likelihood and Inferential Questions
​Finally, we shall link the data with the models, this part includes model asessment, model validation and model selection. We shall see how the link will allow us to answer the inferential questions that gave birth to the experimental study (testing, predicting, forecasting, learning,...) and we will mention diferent approaches to give an answer from the statistical point of view (frequentist statistics, Bayesian statistics, data-based statistics).

Organizers
CRM

Speaker
Massimo Stella, University of Southampton, UK

Description

Language is a complex system, with a hierarchical set of units interacting on several levels. At the level of individual words, psycholinguists conjecture that the interactions among words are encoded within the human mind in the so-called human mental lexicon (HML), i.e. a mental dictionary where words are stored together with their linguistic data. Empirical research has shown that the interaction patterns among words have an impact in learning, storing and retrieving words from the HML, hence the importance of considering the structure of such relationships through a network paradigm.

In [1, 2] we proposed a series of quantitative null models for phonological networks (PNs), where nodes represent words and links represent phonological similarities (i.e. two phonetic transcriptions having edit distance one). Our null models, based on site percolation and Markov processes, suggest the presence of additional constraints in the assembly of real words, such as (i) the avoidance of large degrees, (ii) and the avoidance of triadic closure, which are both compatible with previous empirical findings about avoiding word confusability.

In [3] we extended previous analyses of the HML by adopting a multiplex network framework, including (i) phonological similarities, (ii) synonyms and (iii) free associations. This multi-layered structure was used for investigating how phonological and semantic relationships influence word acquisition through a toy model of lexicon growth driven by the phonological level. When similar sounding words are preferentially learned, the lexicon grows according to the local structure of the whole multiplex, otherwise features of the semantic layers and frequency become predominant, instead.

In [4] we introduced the framework of multiplex lexical networks, i.e. multiplex networks where nodes represent words connected differently on different layers but also endowed with exogenous features such as frequency and age of acquisition. We adopted the multiplex lexical networks of English toddlers for investigating and predicting their normative learning of words between age 18 and 30. Multiplex measures resulted into a higher predictive power of acquired words when compared to single-layer network measures or exogenous estimators such as frequency.

[1] M. Stella and M. Brede, Patterns in the English language: phonological networks, percolation and assembly models, JSTAT, P05006 (2015).
[2] M. Stella and M. Brede, Investigating the Phonetic Organisation of the English Language via Phonological Networks, Percolation and Markov Models, accepted in Proceedings of ECCS2014, Lecture Notes in Computer Science, Springer (2015).
[3] M. Stella and M. Brede, Mental Lexicon Growth Modelling Reveals the Multiplexity of the English Language, Proceedings of the 7th Workshop on Complex Networks, Springer (2016).
[4] M. Stella, N. Beckage and M. Brede, Multiplex lexical networks reveal patterns in early word acquisition in children, https://arxiv.org/abs/1609.03207 (2016).